Improved DWT Algorithm for Filtering of MRI Images for an Efficient Diagnosis

Abstract

Medical scan images allow the Expert to identify the abnormal regions present in the image at its earlier stage. The detection of irregularity in the images becomes a difficult task; if the noise captures under any noisy effect. Such a difficulty is overwhelmed by image filtering after the scanning process. The methodology first converts the RGB to a grayscale image. After that, image filtering has done by the modified DWT (Discrete Wavelet Transform) filtering approach. To evaluate the performance of the proposed algorithm, it utilizes variegated performance metrics includes PSNR (Peak Signal to Noise Ratio), SSIM (Structural Similarity Index Measure), NK (Normalized Cross-Correlation), SC (Structural Content), MD (Maximum Difference), and NAE (Normalized Absolute Error). It has computed between the denoised image and the input image. The experimentation has to be computed on four different medical images, which incorporates brain tumor, skin lesion, chest X-Ray, and lung cancer. The average attained for PSNR, SSIM, NK, SC, and NAE were better for overall set of images. These results exploit that the new approach can potentially improve the filtering performance.